[LCWO LOGO]  

Login

User name:
Password:


Language
Български Português brasileiro
Bosanski Català
繁體中文 Česky
Dansk Deutsch
English Español
Suomi Français
Ελληνικά Hrvatski
Magyar Italiano
日本語 Bahasa Melayu
Nederlands Norsk
Polski Português
Română Русский
සිංහල Slovenščina
Srpski Svenska
ภาษาไทย Türkçe
Українська 简体中文

Who is online? (16)


LCWO Discussion Forum [Atom LCWO Forum Feed]

This is a simple discussion forum for LCWO users. Feel free to use it for any kind of discussion related to this website.

Thread: Levenshtein-Errors

Back to the Forum

AuthorText


Posted: 2008-10-23 18:29
Hi,

What are Levenshtein-Errors? Even when I get everything correct, I get a Levenshtein-Errors: 1 = 1.6 %

The Wikipedia article doesn't really help :(

Many thanks,

LT
Administrator


Posted: 2008-10-23 18:40
In simple terms, it'd be the number of operations (insert, delete) you need to correct the entered text to the sent text. It'll even count one error on a superfluous space at the end of the message.

However, only the lower of the two calculated errors is counted.

I added the second way to calculate the errors/accuracy to avoid that people who forget a space between two groups get a very low score by the errors on _all_ subsequent groups (since the first method is splitting by groups).

It's all not quite perfect yet :)


Posted: 2008-10-24 11:30
Many thanks!


Posted: 2008-10-24 13:46
Just to avoid confusion, could you add something like the following?

Errors:


Posted: 2008-10-24 14:15
Sorry the PHP did not display.
Administrator


Posted: 2008-10-24 15:04
striptags() works :)

Guess I will just display the better result of the two, and if the normal errors are much higher than the Levenshtein errors, display a hint that probably the spaces between the groups are missing.


Posted: 2008-10-27 08:53
Fabian,

personally I think the wikipedia article is ok. No errors in copying leads to 0 faults - at least for me.
I like to see both the absolute and Levenstein errors - just as it is implemented now. It is then up to the user in which error counter he is more interested.

Regards,

Gerd.


Posted: 2008-10-31 06:40
I missed more than half, and the levenshtein accuracy was over 90%. Levenshtein is not a good method of determining accuracy.


Posted: 2008-10-31 06:42
It should count spaces as characters, and not count all subsequent characters as errors if one is missed. Plus, on my slow connection, there's a BIG LONG PAUSE after the first character, then the rest seem normal.


Posted: 2008-11-02 14:30
This would probably mean another value in a database and would probably be too much work - but could it be user defined whether they wanted to see the higher of the two accuracies or see both.

Could it be defined during the signup process or just in change CW settings or edit account?

Many thanks.


Posted: 2008-11-03 08:35
qwasty,

if you miss more than 90% than you probably need no one else to count errors for you...

It is also helpful after checking results to use browser back button, insert the occassionally missing space manually and check again.

Regards,

Gerd

Back to the Forum

You must be logged in to post a message.

$Id: forum.php 62 2015-01-12 17:34:44Z fabian $